Effects of Stimulant Medication on Growth Rates

SPECIAL SECTION

Effects of Stimulant Medication on Growth Rates Across 3 Years in the MTA Follow-up JAMES M. SWANSON, PH.D., GLEN R. ELLIOTT, PH.D., M.D., LAURENCE L. GREENHILL, M.D., TIMOTHY WIGAL, PH.D., L. EUGENE ARNOLD, M.D., BENEDETTO VITIELLO, M.D., LILY HECHTMAN, M.D., JEFFERY N. EPSTEIN, PH.D., WILLIAM E. PELHAM, PH.D., HOWARD B. ABIKOFF, PH.D., JEFFREY H. NEWCORN, M.D., BROOKE S.G. MOLINA, PH.D., STEPHEN P. HINSHAW, PH.D., KAREN C. WELLS, PH.D., BETSY HOZA, PH.D., PETER S. JENSEN, M.D., ROBERT D. GIBBONS, PH.D., KWAN HUR, PH.D., ANNAMARIE STEHLI, M.P.H., MARK DAVIES, M.S., JOHN S. MARCH, M.D., M.P.H., C. KEITH CONNERS, PH.D., MARK CARON, PH.D., AND NORA D. VOLKOW, M.D.

ABSTRACT Objective: To evaluate the hypothesis of stimulant medication effect on physical growth in the follow-up phase of the Multimodal Treatment Study of Children With ADHD. Method: Naturalistic subgroups were established based on patterns of treatment with stimulant medication at baseline, 14-, 24-, and 36-month assessments: not medicated (n = 65), newly medicated (n = 88), consistently medicated (n = 70), and inconsistently medicated (n = 147). Analysis of variance was used to evaluate effects of subgroup and assessment time on measures of relative size (z scores) obtained from growth norms. Results: The subgroup ! assessment time interaction was significant for z height (p < .005) and z weight (p < .0001), due

primarily to divergence of the newly medicated and the not medicated subgroups. These initially stimulant-naı¨ve subgroups

had z scores significantly 90 at baseline. The newly medicated subgroup showed decreases in relative size that reached asymptotes by the 36-month assessment, when this group showed average growth of 2.0 cm and 2.7 kg less than the not medicated subgroup, which showed slight increases in relative size. Conclusions: Stimulant-naı¨ve school-age children with Combined type attention-deficit/hyperactivity disorder were, as a group, larger than expected from norms before treatment but show stimulant-related decreases in growth rates after initiation of treatment, which appeared to reach asymptotes within 3 years without evidence of growth rebound. J. Am. Acad. Child Adolesc. Psychiatry, 2007;46(8):1014Y1026. Key Words: attention-deficit/hyperactivity disorder, growth, methylphenidate, side effects, long-term outcome.

Accepted January 8, 2007. Please see end of text for author affiliations. The work reported was supported by cooperative agreement grants and contracts from the National Institute of Mental Health to the following: University of California, Berkeley: U01 MH50461 and N01MH12009; Duke University: U01 MH50477 and N01MH12012; University of California, Irvine: U01 MH50440 and N01MH 12011; Research Foundation for Mental Hygiene (New York State Psychiatric Institute/Columbia University): U01 MH50467. The opinions and assertions contained in this report are the private views of the authors and are not to be construed as official or as reflecting the views of the National Institute of Mental Health, the National Institutes of Health, or the Department of Health and Human Services. Correspondence to Dr. James M. Swanson, UCI Child Development Center, 19722 MacArthur Blvd., Irvine, CA 92612; e-mail: [email protected]. 8567/07/4608-1014!2007 by the American Academy of Child and Adolescent Psychiatry. DOI:10.1097/chi.0b013e3180686d7e

1014

The use of stimulant medications (see American Academy of Child and Adolescent Psychiatry, 2002) to treat children with attention-deficit/hyperactivity disorder (ADHD) has increased dramatically since 1990 (Olfson et al., 2002; Swanson et al., 1995), despite limited information on long-term effects (Charach et al., 2004; Gillberg et al., 1997). The Multimodal Treatment Study of Children With ADHD (MTA) was intended to fill this gap (see Arnold et al., 1997). The primary reports of outcome in the MTA focused on efficacy, with assessment of outcome at the end of the initial 14-month treatment phase (MTA Cooperative Group, 1999) and at the first follow-up 10 months later, 24 months after initiation

J. AM. ACAD. CHILD ADOLESC. PSYCHIATRY, 46:8, AUGUST 2007

EFFECTS OF STIMULANT MEDICATION ON GROWTH

of treatment (MTA Cooperative Group, 2004a). The companion reports in this special section (Jensen et al., 2007; Molina et al., 2007; Swanson et al., 2007) focus on efficacy and changes in effectiveness at the end of the second follow-up phase, 36 months after initiation of treatment. The secondary analyses reported here focus on a possible side effect of stimulant medication on growth rates of school-age children with ADHD. The hypothesis of a stimulant-related decrease in physical growth rates (a slowing in the gain of height and weight) was proposed by Safer et al. (1972). Consensus reviews and statements decades apart (see NIH Consensus Committee, 1998; Roche et al., 1979) discounted the clinical significance of this hypothesis, based on prevailing views that short-term growth suppression may occur but does not result in longterm effects on ultimate size. Two hypotheses that offer explanations for this discrepancy of short- and longterm effects on growth and relative size are the growth rebound hypothesis (Safer et al., 1975) and the delayed maturation hypothesis (Spencer et al., 1996). The evidence of growth rebound is inconsistent and mostly circumstantial. Safer et al. (1975) reported that when children were treated with stimulant medications (D-amphetamine or methylphenidate) during the school year, then growth rates for both height and weight were less than expected based on norms, but if medication was discontinued during the summer, growth rates then appeared to be greater than expected, suggesting growth rebound. Satterfield et al. (1979) reported that growth suppression was manifested in the first year of treatment but was followed by growth rebound (greater than expected growth) whether or not stimulant medication was discontinued in the summer. However, multiple studies have not replicated this effect of growth rebound during continued treatment with stimulant medication. Mattes and Gittelman (1983) evaluated groups of children with continuous treatment for 1 to 4 years and documented stimulant-related growth deficit that accumulated with the duration of treatment. It was still present during the final year of observation in the group treated for 4 years with methylphenidate, suggesting that growth slowing did not abate during continuous treatment. Apparently, the growth rebound hypothesis gained prominence and acceptance based on long-term followup studies that suggested that ultimate height in

J. AM. ACAD. CHILD ADOLE SC . PSYC HIATRY, 46: 8, AUGUST 2007

adulthood was not affected by childhood treatment with stimulants. Klein and Mannuzza (1988) followed the Mattes and Gittelman (1983) sample into adulthood and compared their average size to a nonclinical control group. No difference in height or weight was observed. Kramer et al. (2000) followed the Loney et al. (1981) sample into adulthood and compared height and weight to control groups drawn from family members (unmedicated brothers and fathers), the community (randomly selected classmates), and clinical groups (individuals with clinical problems but never medicated). No differences in self-reported height or weight were documented. Because ultimate height appeared not to be compromised in adults who had shown growth slowing during treatment as children, the implication of these studies was that growth rebound must have occurred. However, growth rebound was not measured directly in these influential studies. Spencer et al. (1996) proposed that maturational lag characterized the clinical condition of ADHD, which produced a reduction in growth rate that was correlated with but not necessarily a result of treatment with stimulant medication. This hypothesis was based on data from a 4-year follow-up of a longitudinal study of a group of 124 boys with ADHD, all but 10 with a lifetime history of treatment with medication, compared to a control group of 109 boys without ADHD. However, growth measures were obtained only at one point in time in the longitudinal study, providing only cross-sectional evaluation of size. The key findings of the study were that height and weight deficits were evident in the younger but not the older adolescent children with ADHD and that there was no association between height deficits and treatment with various drug classes (stimulant or other), various dose regimens (robust or mild), and duration of treatment. Based on this, they proposed that slow growth in ADHD children was due primarily to early but temporary manifestations of the ADHD itself and may be disorder related rather than treatment related. Using the same design, Biederman et al. (2003) described the growth of a group of 140 girls with ADHD, most (70%) with a lifetime history of treatment with medication, compared with a control group of 122 non-ADHD girls. They reported a similar pattern as Spencer et al. (1996) reported for boys: the group of preadolescent but not adolescent girls was shorter than the control group, but this difference was not significant for relative size

1015

SWANSON ET AL.

adjusted for age and was not related to medication history. Comparisons of the two samples showed no significant effect of sex. Based on this, they concluded that treatment with stimulant medication does not have an adverse impact on growth of children with ADHD. In recent studies of possible effects of stimulant medication on growth, the findings are inconsistent. Two recent chart-review studies of children treated with stimulant medication in clinical practices in the United States (Lisska and Rivkees, 2003) and Australia (Poulton and Cowell, 2003) have shown reductions in growth rates for up to 3 years without evidence of growth rebound. In two other recent chart-review studies, no evidence was found to support the hypothesis that clinical treatment with stimulant medications reduced the growth rates of children with ADHD (Pliszka et al., 2006; Spencer et al., 2006). However, in both of these studies, the treatment regimens allowed for medication holidays, so they did not evaluate the effects of continuous treatment. A recent prospective study of the initiation of methylphenidate treatment in preschool children, the Preschool ADHD Treatment Study, did evaluate effects of stimulant medication on growth when treatment was monitored frequently and was maintained for the initial period of 1 year, and under these conditions, significant reductions in growth rates were documented for both height and weight (Swanson et al., 2006). When the MTA was initiated (see Arnold et al., 1997) and when the initial analyses were performed (see MTA Cooperative Group, 1999), there was a strong consensus (NIH Consensus Committee, 1998; Roche et al., 1979) that growth suppression either was clinically insignificant (due to a temporary effect of initial treatment with stimulant medication followed by growth rebound) or was an artifact of a characteristic of the clinical condition that resulted in slow growth that was correlated with the use of stimulants but not a consequence of this treatment. Based on this consensus, the possible risk of stimulant-related growth suppression was not a consideration (e.g., it was not mentioned in the consent forms for the MTA), and a test of the hypothesis of stimulant-related growth suppression was not addressed as a specific aim of the MTA. It is not surprising, then, that the initial evaluations of outcome emphasized characterization of long-term efficacy (MTA Cooperative Group, 1999) and that the evaluation of the measures of height and weight were not

1016

included in the initial report of the effects of treatment. The first evaluation of growth in the MTA was part of a report of outcome at the first follow-up 2 years after initiation of treatment (MTA Cooperative Group, 2004b). In this report the intent-to-treat evaluation of assigned treatment revealed that at 14 months the two groups assigned to systematic stimulant medication showed an initial reduction in growth rate for height (j1.23 cm/year) and weight (j2.48 kg/year), which dissipated by the 24-month assessment. However, this dissipation was apparently due to differences between assigned and actual treatment. Changes in height and weight in naturalistic subgroups defined by consistent treatment (i.e., with stimulant medication at both the 14- and 24-month assessments or no treatment at either assessment) revealed continued stimulant-related reductions in annual growth rates in the second year for height (j1.04 cm/year) and weight (j1.22 kg/year). However, in the 24-month evaluation of growth in the MTA, we did not evaluate growth in terms of normed relative size (z scores) or the effect of previous treatment with stimulant medication on initial size upon entry into the MTA. Here we focus on the evaluation of the hypothesis of long-term growth slowing, this time with more rigorous methods and a longer follow-up period than in the previous report (MTA Cooperative Group, 2004b). METHOD The characteristics of the clinical sample and the general methods of the MTA are described in previous publications (MTA Cooperative Group, 1999, 2004a) and in one of the companion papers in this issue (Jensen et al., 2007), so only a brief description is provided here. For the MTA, 579 children between 7.0 and 9.9 years of age (80% males, 61% white, 20% African American, and 8% Hispanic) with confirmed diagnosis of ADHD Combined type were recruited from a variety of sources at seven sites in North America and randomly assigned to one of four treatment conditions: medication management (MedMgt), behavior therapy (Beh), the combination of these two modalities (Comb), and usual community care (CC). The MTA Medication Algorithm was used to manage stimulant medications for most of the participants in the MedMgt and Comb groups over the initial 14-month treatment phase (Greenhill et al., 2001; Vitiello et al., 2001), and some participants in the CC (67%) and Beh (26%) received stimulant medication from clinicians in the community. Compliance with initial randomly assigned treatment conditions was not perfect during the treatment phase, and adherence to these conditions waned during the follow-up phases of the MTA, which made the evaluation of actual as well as assigned treatment advisable (MTA Cooperative Group, 2004b). A record of actual treatment was provided by parental reports on the Services for Children and AdolescentsYParent Interview (SCA-PI; Hoagwood et al., 2004;



TABLE 1 Naturalistic Subgroups Based on SCA-PI Reports of Use of Stimulant Medication Previous 14 Mo 24 Mo Not (n = 65) Newly (n = 88)a Consistently (n = 70)b Inconsistently (n = 147) (n = 13) (n = 18) (n = 2) (n = 9) (n = 23)b (n = 6)b (n = 5) (n = 6)b (n = 18) (n = 19) (n = 11) (n = 8) (n = 9)

No Med No Med Med 58:89 Med No Med No Med No Med No Med Med Med Med No Med No Med Med Med Med

No Med Med Med 80:67 No Med Med No Med No Med No Med Med No Med No Med Med Med Med Med No Med

No Med Med Med 68:79 No Med No Med Med No Med Med No Med Med No Med Med No Med Med No Med Med

36 Mo No Med Med Med 74:73 No Med No Med No Med Med Med No Med No Med Med Med Med No Med Med Med

Note: SCAPI = Services for Children and Adolescents-Parent Interview. a One missing value for weight at baseline assessment. b One missing value for weight at 36-mo assessment.

Jensen et al., 2004) at each assessment, which provided information to estimate how many days stimulant medication was used since the previous assessment and the doses administered, which were expressed in methylphenidate equivalents for amphetamine (! 2) and pemoline (" 5), which were the other stimulants that were available and in use during this trial. From this record, we determined the total number of days treated and total cumulative dose consumed during the study, as well as whether medication was administered during the 30 days before each assessment. In all three subgroups treated with medication, the primary drug used was methylphenidate at all assessment points, with the percentage ranging from 95.4% to 73.5%. At the baseline and 14-, 24-, and 36month assessment points, respectively, the percentages of medicated children taking methylphenidate were the following: consistently (85.4%, 79.7%, 76.8%, and 73.5%; newly (not available, 95.4%, 93.0%, and 85.9%), and inconsistently (94.9%, 93.1%, 92.4%, and 85.1%). In other publications on the MTA outcomes, two definitions of positive medication status have been used. In the secondary analysis of changes in effectiveness and growth at the 24-month assessment, medication use during the previous 30 days (MTA Cooperative Group, 2004b) was the criterion for positive medication status. In the assessment of effectiveness at the 36-month assessment, medication use more than 50% of the days since the previous assessment was the criterion for positive medication use (Jensen et al., 2007; Swanson et al., 2007). To maintain consistency with the previous article on growth (MTA Cooperative Group, 2004b), here we used the 30-day definition; therefore, if medication was used within a 30-day period before the assessment medication status was positive (Med) and otherwise negative (No Med). If an individual_s medication status changed at any assessment point, then they were placed in the inconsistently medicated group. For this binary designation of medication status at four assessment


points, there are 24 patterns of medication use that define 16 possible naturalistic subgroups at the 36-month assessment. The most recent growth norms provided by the U.S. Centers for Disease Control and Prevention (Kuczmarski et al., 2000) were used to transform the raw scores (centimeters and kilograms) into standard scores (z height and z weight). Based on an assumption of normal size at entry to the MTA, the baseline z scores are expected to be near zero, and based on the assumption of normal growth rates during the MTA implementation of the MTA protocol, the z scores for height and weight are expected to remain constant over time. A two-way analysis of variance, with a significance level of p < .05, was used to evaluate the effects of the between group factor (naturalistic subgroups), the within group factor (assessment time), and the interaction on measures of z height and z weight. RESULTS

Complete information was obtained at each time point for 370 of the 485 children in the MTA followup (Jensen et al., 2007). These cases were spread across the 16 possible naturalistic subgroups defined by the patterns of medication status (Med or No Med) across the four assessment points (Table 1), but were concentrated in three of these subgroups that represented consistent patterns over time: not medicated (65 stimulant-naı¨ve children at entry who were not receiving stimulant medication at any assessment point), newly medicated (88 children stimulant naı¨ve at entry who started stimulant medication in the MTA

1017

SWANSON ET AL.

and were receiving medication at each assessment point after baseline), and consistently medicated (70 children receiving stimulant medication during the 30-day period before entry, although the length of the previous treatment was not determined, and at each assessment point). The other subgroups were collapsed into a single group to form a fourth subgroup who were inconsistently medicated (147 children, with reports of medication at some but not all of the assessment points, with ratios of Med to No Med of 58:89, 80:67, 68:79, and 74:73, respectively, across the four assessment points). At the 24-month and 36-month assessment points, we also obtained height and weight measures from 213 of the 279 classmates of the ADHD cases who were recruited from the local normative comparison group (LNCG), which became part of the MTA followup at the 24-month assessment point. Both randomization and self-selection operated to form these naturalistic subgroups. To check whether they differed on some relevant factors at baseline, we used the MTA database to obtain information on 26 child, parent, pregnancy, birth, and infant variables. We used one-way analysis of variance to compare the four naturalistic subgroups on each of these variables. As would be expected by chance for 26 F tests, each with 3 and 366 df and an unadjusted significance level of p < .05, 1 (pregnancy length) showed a small but statistically significant between-group difference. After adjustment for multiple tests (0.05/26 = 0.002), none of these tests were statistically significant. It is notable that the four subgroups did not differ on initial size at birth (birth weight), age, parent or teacher ratings of ADHD

symptoms, sex, expected adult size (mid-parent size), welfare status, or maternal smoking (Table 2). The analyses of variance of our measures of physical size produced significant interactions of naturalistic subgroup ! assessment time (Fig. 1 and Table 3) for z height (F 9,1083 = 3.88; p < .005) and z weight (F 9,1,068 = 8.88; p < .0001). The overall interaction was decomposed into two orthogonal components to evaluate the impact of starting medication (the not medicated versus the newly medicated subgroups) and consistency of medication (the consistently medicated versus the inconsistently medicated subgroups). To complete the set of three orthogonal comparisons, the third was derived from a combination of the conditions (Not + Newly vs. Consistently + Inconsistently). This decomposition of variance into nonoverlapping components revealed that the overall interaction was due primarily to the starting medication component, which accounted for 92% of the interaction for height (F3,441 = 11.77; p < .0001) and 72.8% for weight (F3,438 = 21.76; p < .0001). The profiles shown by dashed lines in Figure 1A (for height) and B (for weight) suggest that this effect was due to the significant divergence of the unmedicated clinical control group (the not medicated subgroup) and prospectively medicated group (the newly medicated subgroup). The consistency of medication contrast was also significant, but this was due to a main effect of group without an interaction with assessment time. At the baseline assessment, the consistently medicated subgroup was smaller than the inconsistently medicated subgroup, and this difference was maintained over time,

TABLE 2 Baseline Demographic and Clinical Variables Across the Four Naturalistic Medication Subgroups ANOVA With 3 df (Group) and 369 df (Error) Not Newly Inconsistently Consistently Child variables Birth weight, kg Age at baseline, y SNAP parent compositea SNAP teacher compositea Sex, % male Parent variables Parent average height z score % Welfare % Mother smoked during pregnancy

p

3.42 7.94 1.91 1.87 74

3.35 7.75 1.92 1.89 74

3.30 7.91 2.00 1.91 78

3.46 8.04 2.04 2.08 90

.0990 .8449 .5060 .4840 .0580

0.17 15.38 18.52

0.28 13.64 32.10

0.20 17.81 33.33

0.24 8.57 16.05

.6155 .3426 .2362

Note: ANOVA = analysis of variance; SNAP = Swanson, Nolan, and Pelham. a SNAP composites were ratings in unmedicated state of 18 DSM-IV attention-deficit/hyperactivity disorder symptoms on 0Y3 metric.

1018



as shown by the parallel profiles for these two subgroups for both height and weight (see solid lines in Fig. 1A, B). In Figure 1C the average SNAP scores for the four naturalistic subgroups are presented to provide information on the effectiveness of treatment along with the information on effects related to growth presented in Figure 1A (for height) and B (for weight). In the companion articles (Jensen et al., 2007; Swanson et al., 2007), the effects of actual as well as assigned medication status over time were addressed in detail, with different methods (e.g., time-varying covariates), and at the 36-month assessment, there were no group differences apparent. The analysis of this measure of effectiveness for the naturalistic subgroups confirms the results reported in the companion articles.

Using information from the SCAPI, we estimated the number of days treated with stimulant medication and the ending milligrams/kilogram dose of stimulant medication (in methylphenidate equivalents) for each child. Regression analyses were used to relate these measures to relative size at the 36-month assessment. The fitted equations had negative slopes indicating significant exposure-related reduction in relative height (z height 0.589 Y 0.218[years of treatment], F1,369 = 19.18; p 9 .0001) and relative weight (z weight 0.776 Y 0.183[years of treatment], F1,365 = 12.68; p < .0004) and significant dose-related reduction in relative height (z height 0.493 j 0.0117[mg/kg], F1,365 = 11.04; p .0010) but not relative weight (z weight 0.547 j 0.0045[mg/kg], F1,365 = 1.34; p 9 .2486).

Fig. 1 Profiles for the naturalistic subgroup ! assessment time interaction, showing standardized measures of size (z height [A] and z weight [B]) and efficacy (average SNAP ratings [C]) for the four subgroups formed on the basis of history of medication use before entry into the MTA (baseline), at the end of the MTA treatment phase (14 months), at the first follow-up (24 months) and at the second follow-up (36 months). LNCG = local normative comparison group; SNAP = Swanson, Nolan, and Pelham.


1019

SWANSON ET AL.

TABLE 3 Means, SDs, and SEs for the Naturalistic Subgroups Base 14 Mo 24 Mo Height z score Not med Newly Incons Consis LNCG Weight z score Not med Newly Incons Consis LNCG SNAP Not med Newly Incons Consis LNCG

No.

Mean

SD

SE

Mean

SD

SE

Mean

SD

65 88 147 70 260

0.357 0.178 0.239 j0.140

0.892 1.12 1.02 0.905

0.111 0.119 0.084 0.108

0.415 0.072 0.248 0.165

0.870 1.08 1.02 0.905

0.108 0.491 0.115 0.0435 0.084 0.282 0.108 j0.159 0.229

0.913 1.11 1.12 1.04 0.945

65 88 147 70 259

0.617 0.616 0.430 0.312

0.915 0.999 1.05 0.999

0.113 0.106 0.087 0.119

0.621 0.113 0.311 0.037

0.954 1.14 1.16 1.05

0.118 0.122 0.096 0.125

0.652 0.181 0.35 0.071 0.427

39 63 106 55 251

1.66 1.87 1.76 1.92

0.406 0.424 0.398 0.405

0.050 0.045 0.033 0.048

1.04 0.811 1.10 1.04

0.565 0.509 0.524 0.410

0.070 0.054 0.043 0.049

1.08 1.02 1.16 1.19 0.444

36 Mo SE

Mean

SD

SE

0.113 0.541 0.118 0.0725 0.092 0.310 0.124 j0.12 0.014 0.234

0.894 1.09 1.05 0.894 0.924

0.111 0.116 0.087 0.107 0.015

0.975 1.09 1.21 1.03 1.04

0.121 0.116 0.100 0.123 0.026

0.756 0.293 0.519 0.156 0.425

1.01 1.15 1.25 1.200 1.07

0.125 0.123 0.103 0.143 0.026

0.508 0.521 0.497 0.427 0.945

0.063 0.056 0.041 0.051 0.028

1.01 1.12 1.09 1.10 0.38

0.545 0.555 0.516 0.382 0.924

0.068 0.059 0.043 0.046 0.024

Note: Not med = not medicated; LNCG = local normative comparison group; Newly = newly medicated; Incons = inconsistently medicated; Consis = consistently medicated; SNAP = Swanson, Nolan, and Pelham Rating Scale.

At the 36-month assessment, the average relative size (z height and z weight) of the naturalistic subgroups were negatively related to the average cumulative exposure to stimulant medication. As shown in Figure 2 the totals for the naturalistic subgroups, which were formed using the 30-day criteria at each assessment point, varied as expected: not medicated (28 days and

529 mg), inconsistently medicated (603 days and 16,183 mg), newly medicated (983 days and 31,797 mg), and consistently medicated (1,022 days and 33,576 mg). The stimulant-untreated clinical control group was taller and heavier than average (not medicated: z height 0.541 and z weight 0.765), whereas the two groups that were medicated for about 1,000 days and received

Fig. 2 Size (z height and z weight) at the 36-month assessment time point for the four naturalistic subgroups that differed in exposure (total milligrams and total number of days) to stimulant medication in the MTA. LNCG = local normative comparison group.

1020



930,000 mg methylphenidate over the 3 years of the MTA follow-up were just slightly above or below expected relative size based on norms (newly medicated: z height 0.073 and z weight 0.293; consistently medicated: z height j0.120 and z weight 0.156) and well below the average relative height and weight of the LNCG. The average population SDs for 7- to 12-year-old children, 6.5 cm for height and 5.5 kg for weight, were used to transform the average z scores at the 36month assessment into absolute units (centimeters and kilograms) for comparisons to the stimulantuntreated clinical control group as well as the classmate nonclinical control group. These comparisons revealed that the newly medicated subgroup (which had been only 1.1 cm shorter than the not medicated group at baseline) was at 36 months 3.04 cm shorter and 2.71 kg lighter than the not medicated subgroup and 1.1 cm shorter and 0.7 kg lighter than the LNCG. The consistently medicated subgroup was 2.3 cm shorter and 1.5 kg lighter than the LNCG and 4.21 cm shorter and 3.51 kg lighter than the not medicated subgroup. DISCUSSION

The MTA provides an excellent opportunity to address two key methodological issues in the literature that were identified by Spencer et al. (1996), who recommended Bto differentiate disorder from treatmentrelated growth effects, studies must compare treated children with ADHD with untreated children, and not with unaffected control subjects,[ and that this Bshould be evaluated in future longitudinal studies.[ In this prospective longitudinal study of school-age children with diagnoses of ADHD Combined type, one of the naturalistic self-selected subgroups (derived largely from those randomly assigned to Beh) provided a stimulant-untreated clinical control group. Compared to this not medicated subgroup, we documented a significant stimulant-related decrease in growth rate in the newly medicated subgroup consisting of initially stimulant-naı¨ve individuals (at least for 30 days before entering the study) for whom treatment was initiated and maintained for 3 years of the MTA protocol. These subgroups did not differ in severity of ADHD symptoms or initial relative size at the baseline assessment. The comparison of these two subgroups


in terms of growth velocity after initiating medication revealed the time course of stimulant-related reduction in annual growth rate, which was maximal in the first year, decreased in the second year, and absent in the third year of treatment. We did not document a decrease in relative size in the group of participants with a history of treatment before entry into the MTA protocol during subsequent treatment with stimulant medication over the 3 years. However, this group (the consistently medicated naturalistic subgroup) was smaller than the stimulantnaı¨ve group (the newly medicated naturalistic subgroup) at entry, suggesting that early treatment of children (before the ages of 7Y9 years) with stimulant medication may have produced a reduction of growth rate before entry into the MTA protocol. Of course, other possibilities exist, such as selective early treatment of individuals who are smaller than normal. Similarly, the participants without continuous treatment with stimulant medication (the inconsistently medicated subgroup) did not manifest a decrease in relative size during the MTA protocol but was slightly smaller at baseline and did not increase in relative size as much as the not medicated group over the 3-year follow-up. By the 36-month assessment, this subgroup was significantly smaller than the stimulant-untreated clinical control group (not medicated), even though it was nominally larger than the LNCG. The definition of inconsistent treatment here was not based on a systematic variation of planned medication holidays on the weekend or in summer as in some previous studies (Gittelman-Klein et al., 1988; Safer et al., 1975; Satterfield et al., 1979). Our findings do not support the hypothesis of growth rebound when treatment is continued, as suggested by Satterfield et al. (1979). The newly medicated subgroup manifested a slight increase in relative size from the 24- to 36-month assessment. This suggests an increase in growth rate to a level that is greater than expected based on population norms, which may be interpreted as growth rebound. However, the stimulant-untreated clinical control (the not medicated subgroup) also showed a similar increase in relative size across this time frame. To support the hypothesis of growth rebound, the growth rates of the stimulant-treated groups would have to exceed that of the stimulant-untreated clinical control group, which did not occur in this study. We did not include a

1021

SWANSON ET AL.

randomized withdrawal of medication, which would have provided a test of the growth rebound hypothesis (greater than normal growth velocity for a time after medication is stopped) that was suggested by Safer et al. (1975). Similarly, our findings do not support the hypothesis of delayed maturation in children with ADHD. At entry into the MTA, the height and weight of stimulant-naı¨ve 7.0- to 9.9-year-old children were greater than expected (rather than smaller as would be predicted by delayed maturation). For those left untreated with stimulant medication, the growth rate over the next 3 years was greater as opposed to less than that of the LNCG (the classmate control group) or the population norms. This is an important new finding from the MTA, suggesting a disorder-related accelerated rather than delayed maturation. Basic science research collaborators with expertise on the neurotransmitter dopamine in multiple pathways at multiple levels of analysis (Bosse et al., 1997; Posner and Raichle, 1994; Volkow et al., 2002) suggested how the mechanism of action of stimulant medications may affect growth. Studies using positron emission tomography show that clinical doses of methylphenidate in adults (Volkow et al., 2002) and adolescents (Neto et al., 2002) increased (rather than decreased, as proposed by some theories) extrasynaptic levels of dopamine in the striatum due to potent blockade of dopamine transporters, but with a time course that minimized the reinforcing (euphoric) effects accompanying intravenous doses (Volkow et al., 1999). Studies of mice (Bosse et al., 1997) that lack the dopamine transporter show that high levels of dopamine occur in the hypothalamus as well as the striatum. The excess dopamine is dispersed by blood flow, reaches the pituitary, and retards growth in these animals. Caron (2004) summarized this converging information from basic science studies and speculated that a common synaptic mechanism (dopamine transporter blockade and increased dopamine levels) in different brain regions (hypothalamus and striatum) may mediate side effects (reduction in growth rate) as well as efficacy (symptom reduction). Limitations

The most important limitation of this study is the lack of random assignment to the subgroups that were

1022

evaluated here. Instead, these naturalistic subgroups were formed partially by the assignment to the initial treatment conditions provided in the MTA protocol with (MedMgt and Comb) and without (Beh) stimulant medication, and then by choices over time to start, stop, or continue the use of stimulant medication. These decisions depended on the assigned treatment group. Unless the assignment to a condition that included medication was refused, the only way to be in the Bnever medicated[ group was to enter stimulant naı¨ve and be assigned to Beh or be in the minority of CC children who were not medicated. Thus, the never-medicated group was determined more by original randomization than the other three naturalistic groups, which were compatible with assignment to any of the four groups (Comb, MedMgt, CC, even Beh if nonprotocol medication was taken before 14 months). Self-selection for the never medicated group was passive, amounting to not starting medication, whereas for the other three groups, self-selection would require taking action by continuing, starting, or stopping medication. Furthermore, the self-selection to stay not medicated was based largely on satisfactory results with the assigned behavioral treatment, whereas decisions to continue or start medication were not necessarily based on satisfaction with treatment results. At 14 months 26% of Beh subjects had found it necessary to add nonprotocol medication because Beh alone was not sufficient treatment, and more than half of those had taken medication 950% of the time from baseline. These subjects were counted in a medicated group for purpose of the analyses presented here, which may have weeded out the worst treatment responders and more serious cases from the never-medicated group. Thus, it is possible that the never-medicated group was pared down to good responders and the medicated groups enriched with poor Beh responders. These limitations of the naturalistic medication groups must be kept in mind when attempting to interpret Figure 1c, which suggests lack of effectiveness for stimulant medication at the 36-month assessment in the presence of lingering but significant stimulantrelated reduction in growth rate. Another serious limitation is the length of the followup reported here, which covered only 3 years after the initiation of treatment in the MTA protocol. Because the participants in the study entered between the ages of 7 and 9 years, they were between the ages of 10 to



12 years when the 36-month assessments were performed. Subsequent reports of the continued follow-up of the MTA sample will provide assessment after puberty, in adolescence, and eventually in adulthood. These assessments will allow evaluation of the impact of childhood treatment on ultimate size, which is not addressed in this report. A third major limitation of this study is that only one stimulant medication (methylphenidate) and one formulation (immediate release) of stimulant medication were systematically evaluated. The MTA implemented state-of-the-art treatment in the early 1990s and used an immediate-release formulation of methylphenidate (Ritalin, which then was prescribed for most cases when stimulant medication was used to treat children with ADHD (Arnold et al., 1997; Greenhill et al., 2001; MTA Cooperative Group, 2004b). However, in 1998, 4 years after the initiation of the MTA, another stimulant, the 75:25 ratio of D,L - amphetamine (once used for weight control [Obetrol]) was renamed Adderall and was evaluated for the treatment of ADHD. Adderall was shown to be effective in a double-blind study (Swanson et al., 1998) and soon after became widely used in clinical practice. Because Adderall was not available when the MTA was initiated, its effect was not evaluated and thus could not be reported here. Also, after 2000, controlled-release formulations of both methylphenidate (Concerta: Swanson et al., 1999, 2000, 2003; Wolraich et al., 2001; Metadate CD: Greenhill et al., 2002; Swanson et al., 2004; Wigal et al., 2003; Ritalin LA: Biederman et al., 2003) and amphetamine (Adderall SR; Greenhill et al., 2001; McCracken et al., 2003) were developed and became widely used in clinical practice, rapidly replacing the prescription of multiple daily doses of immediaterelease stimulant medications for the treatment of ADHD. Because these controlled-release formulations were not available or used when the MTA was initiated or during the initial two follow-up phases, they were not evaluated and could not be reported here. Thus, the findings here strictly apply to the immediate-release formulation of methylphenidate. A fourth limitation was the definition of medication use based on information from the SCA-PI. Two different definitions of medication use have been considered for the MTA analyses of efficacy (i.e., within 30 days before an assessment and 50% of the days since the previous assessment), and consistency of these


classifications by medication have been considered as well, as in the naturalistic subgroups described here and in previous publications on the findings of the MTA follow-up. Also, other definitions are possible and some have been considered here (i.e., total exposure in terms of milligrams consumed or days treated), but the specific contributions of dose or consistency of medication use have not been adequately evaluated. A fifth limitation was the lack of evaluation of planned medication holidays on the weekends or during the summer, as some guidelines have proposed. For example, the use of summer (or periodic) holidays, which represent relatively long intervals of nontreatment interspersed within blocks of medication use in clinical practice, has been proposed as a strategy to mitigate stimulant-related reductions in growth rates (Safer et al., 1975), but this was not evaluated in the present analyses. In two recent studies (Pliszka et al., 2006; Spencer et al., 2006) in which drug holidays were allowed but were not systematically varied, stimulantrelated reductions in growth rates were not documented. This and other unspecified factors still must be considered to understand the effects of stimulant medication on growth rates. Clinical Implications

The practice parameter of the American Academy of Child and Adolescent Psychiatry (2002) regarding the use of stimulant medication (presently under revision) suggests that height and weight be measured before initiating treatment (p 28S), and for the management of possible treatment-related side effects recommends the regular assessment of weight (but not height) as a Bclinical guideline[ but not as a Bminimal standard[ (p 29S). During the first two stages of treatment to titrate dose and select among alternative stimulants, weekly to monthly monitoring of both height and weight is suggested (Table 2, p 38S), but recommendations for long-term monitoring are not explicitly provided. In the section on Bcomplications and side effects[ (p 44S), the prevailing view is stated, indicating that the literature supports short-term decrements in rate of weight acquisition but not height acquisition, and no long-term effect on ultimate height. The findings reported here from the prospective MTA follow-up suggest that short-term (up to 2 years) effects on height as well as weight may occur when

1023

SWANSON ET AL.

initiating treatment in stimulant-naı¨ve school-age children with ADHD Combined type. This conclusion is based on two opposing patterns of growth in the MTA: greater than expected annual growth rates in the subgroup of ADHD children who were not treated with stimulant medication and smaller than expected annual growth rates in the subgroup of ADHD children who were continuously treated. This difference in annual growth rates may be present for up to 3 years of treatment and accumulate to result in a difference of about 2.0 cm in height and about 2.0 kg in weight. The finding of a reduction in growth rate in height that extends up to 3 years is consistent with two recent retrospective reviews of clinical charts (Lisska and Rivkees, 2003; Poulton and Cowell, 2003), but not with two others (Pliszka et al., 2006; Spencer et al., 2006). Despite this inconsistency across chart-review studies, the finding from the prospective study of schoolage children reported here as well as in a prospective study of preschool children (Swanson et al., 2006) may provide enough evidence to consider revision of clinical practice parameters to acknowledge the possibility of stimulantrelated slowing in the usual developmental gains in height as well as weight during the course of treatment of prepubertal children with ADHD. The Multimodal Treatment Study of Children with ADHD (MTA) was a National Institute of Mental health (NIMH) cooperative agreement randomized clinical trial involving six clinical sites. Collaborators from the National Institute of Mental Health: Peter S. Jensen, M.D. (currently at Columbia University, New York), L. Eugene Arnold, M.D., M.Ed. (currently at Ohio State University), Joanne B. Severe, M.S. (Clinical Trials Operations and Biostatistics Unit, Division of Services and Intervention Research), Benedetto Vitiello, M.D. (Child and Adolescent Treatment and Preventive Interventions Research Branch), Kimberly Hoagwood, Ph.D. (currently at Columbia University); previous contributors from NIMH to the early phase: John Richters, Ph.D. (currently at National Institute of Nursing Research); Donald Vereen, M.D. (currently at National Institute on Drug Abuse). Clinical sites and principal investigators and co-investigators are University of California, Berkeley/San Francisco: Stephen P. Hinshaw, Ph.D. (Berkeley), Glen R. Elliott, M.D., Ph.D. (San Francisco); Duke University: C. Keith Conners, Ph.D., Karen C. Wells, Ph.D., John March, M.D., M.P.H., Jeffrey Epstein, Ph.D.; University of California, Irvine/Los Angeles: James Swanson, Ph.D. (Irvine), Dennis P. Cantwell, M.D. (deceased, Los Angeles), Timothy Wigal, Ph.D. (Irvine), Annamarie Stehli, M.P.H. (Irvine); Long Island Jewish Medical Center/Montreal Children_s Hospital: Howard B. Abikoff, Ph.D. (currently at New York University School of Medicine), Lily Hechtman, M.D. (McGill University, Montreal); New York State Psychiatric Institute/Columbia University/Mount Sinai Medical Center: Laurence L. Greenhill, M.D. (Columbia

1024

University), Jeffrey H. Newcorn, M.D. (Mount Sinai School of Medicine), Mark Davies (New York State Psychiatric Institute); University of Pittsburgh: William E. Pelham, Ph.D. (currently at State University of New York, Buffalo), Betsy Hoza, Ph.D. (currently at University of Vermont, Burlington), Brooke Molina, Ph.D. Original statistical and trial design consultant: Helena C. Kraemer, Ph.D. (Stanford University). Follow-up phase statistical collaborators: Robert D. Gibbons, Ph.D. (University of Illinois, Chicago), Sue Marcus, Ph.D. (Mount Sinai School of Medicine), Kwan Hur, Ph.D. (University of Illinois, Chicago). Collaborator from the Office of Special Education Programs/U.S. Department of Education: Thomas Hanley, Ed.D. Collaborator from Office of Juvenile Justice and Delinquency Prevention/Department of Justice: Karen Stern, Ph.D. Non-MTA collaborators were Marc Caron, Ph.D. (Duke University), and Nora D. Volkow (National Institute of Drug Abuse and Brookhaven National Laboratory).

Disclosure: During the course of the MTA, since 1992: Dr. Swanson has received research support from Alza, Richwood, Shire, Celgene, Novartis, Celltech, Gliatech, Cephalon, Watson, CIBA, Janssen, and McNeil; has been on the advisory board of Alza, Richwood, Shire, Celgene, Novartis, Celltech, UCB, Gliatech, Cephalon, McNeil, and Eli Lilly; has been on the speakers_ bureaus of Alza, Shire, Novartis, Celltech, UCB, Cephalon, CIBA, Janssen, and McNeil; and has consulted to Alza, Richwood, Shire, Celgene, Novartis, Celltech, UCB, Gliatech, Cephalon, Watson, CIBA, Janssen, McNeil, and Eli Lilly. Dr. Elliott has received research funding from Cephalon, McNeil, Shire, Sigma Tau, and Novartis; has consulted to Cephalon and McNeil; and has been on the speakers_ bureaus of Janssen, Eli Lilly, and McNeil. Dr. Greenhill has received research funding or has been on the speakers_ bureaus of Eli Lilly, Alza, Shire, Cephalon, McNeil, Celltech, Novartis, Sanofi Aventis, Otsuka, and Janssen. Dr. Wigal has received research funding from Eli Lilly, Shire, Novartis, and McNeil, and has been on the speakers_ bureaus of McNeil and Shire. Dr. Arnold has received research funding from Celgene, Shire, Noven, Eli Lilly, Targacept, Sigma Tau, and Novartis; has consulted to Shire, Noven, Sigma Tau, Ross, and Organon; and has been on the speakers_ bureaus of Abbott, Shire, McNeil, and Novartis. Dr. Vitiello has consulted to Richwood Pharmaceuticals. Dr. Hechtman has received research funding from the National Institute of Mental Health, Eli Lilly, GlaxoSmithKline, Janssen-Ortho, Purdue Pharma, and Shire; has been on the speakers_ bureaus of the National Institute of Mental Health, Eli Lilly, Janssen-Ortho, and Shire; and has been on the advisory boards of Eli Lilly, Janssen-Ortho, Purdue Pharma, and Shire. Dr. Epstein has received research funding from McNeil, Shire, Eli Lilly, and Novartis; has been on the advisory board of Shire; and has been on the speakers_ bureaus of Shire and McNeil. Dr. Pelham has received research funding from Alza, Shire, Noven, Eli Lilly, and Cephalon; has served on advisory boards or has consulted to Alza/McNeil Richwood/Shire, Noven, Eli Lilly, Cephalon, Novartis, Celgene, and Abbott; and has been on the speakers_ bureaus of Shire, and McNeil. Dr. Abikoff has received research funding from McNeil, Shire, Eli Lilly, and BristolMyers Squibb; has consulted to McNeil, Shire, Eli Lilly, Pfizer, Celltech, Cephalon, and Novartis; and has been on the speakers_ bureaus of McNeil, Shire, and Celltech. Dr. Newcorn has received research funding from or has been on the speakers_ bureaus of Eli Lilly, Alza, Shire, Celgene, McNeil, Celltech/UCB, Novartis, Sanofi Aventis, Janssen, and Bristol-Myers Squibb. Dr. Hinshaw has consulted to Noven and Sigma Tau and has been on the speakers_ bureau of McNeil. Dr. Hoza has received research funding from MediaBalance, Inc., and has received support for educational conferences from Abbott



Laboratories. Dr. Jensen has received research funding from McNeil and unrestricted grants from Pfizer; has consulted to Best Practice, Inc., Shire, Janssen, Novartis, and UCB; and has been on the speakers_ bureaus of Janssen-Ortho, Alza, McNeil, UCB, CME Outfitters, and the Neuroscience Education Institute. Dr. March has been a consultant or scientific advisor to or received research funding from Eli Lilly, Pfizer, Wyeth, Jazz, MedAvante, Shire, Cephalon, Organon, McNeil, and AstraZeneca; serves on a DSMB for Organon, Johnson & Johnson, and AstraZeneca; and holds stock in MedAvante. Dr. Conners has received research funding from Celgene, Shire, Noven, Eli Lilly, Targacept, and Novartis; has consulted to Celgene, Shire, Novartis, Alza, and Noven; is on the Eli Lilly Advisory Committee, and has been on the speakers_ bureaus of Shire, McNeil, and Novartis. Dr. Caron has received research funding from and has consulted to Lundbeck, has been on the Scientific Advisory Board of Acadia Pharmaceutical, and has had an unrestricted grant from Bristol-Myers Squibb. Dr. Volkow has received research funding from GlaxoSmithKline. The other authors have no financial relationships to disclose. REFERENCES American Academy of Child and Adolescent Psychiatry (2002), Practice parameter for the use of stimulant medications in the treatment of children, adolescents, and adults. J Am Acad Child Adolesc Psychiatry 41S:26SY49S Arnold LE, Abikoff HB, Cantwell DP, Conners CK, Elliott GR, Greenhill LL (1997), NIMH collaborative multimodal treatment study of children with ADHD (MTA): design challenges and choices. Arch Gen Psychiatry 54:865Y870 Biederman J, Faraone S, Monuteaux M, Plunkett E, Gifford J, Spencer T (2003), Growth deficits and attention-deficit/hyperactivity disorder revisited: impact of gender, development, and treatment. Pediatrics 111: 1010Y1016 Bosse R, Fumagalli F, Jaber M et al. (1997), Anterior pituitary hypoplasia and dwarfism in mice lacking the dopamine transporter. Neuron 19:127Y138 Caron M (2004), Growth in the dopamine transporter knockout animal model of ADHD. Neuropsychopharmacology 29:S55 Charach A, Ickowicz A, Schachar R (2004), Stimulant treatment over five years: adherence, effectiveness, and adverse effects. J Am Acad Child Adolesc Psychiatry 43:559Y567 Gillberg C, Melander H, von Knorring A et al. (1997), Long-term stimulant treatment of children with attention-deficit hyperactivity disorder. symptoms. A randomized, double-blind, placebo-controlled trial. Arch Gen Psychiatry 54:857Y864 Gittelman-Klein R, Landa B, Mattes JA, Klein DF (1988), Methylphenidate and growth in hyperactive children. Arch Gen Psychiatry 36:212Y217 Greenhill LL, Findling RL, Swanson JM, and the MPH MR ADHD Study Group (2002), A double-blind, placebo-controlled study of modifiedrelease methylphenidate in children with attention-deficit/hyperactivity disorder. Pediatrics 109:E39YE46 Greenhill LL, Swanson JM, Vitiello B et al. (2001), Impairment and deportment responses to different methylphenidate doses in children with ADHD: the MTA titration. J Am Acad Child Adolesc Psychiatry 40:180Y187 Hoagwood K, Jensen P, Arnold LE et al., and the MTA Cooperative Group (2004), Reliability of the Services for Children and Adolescents Parent Interview (SCAPI). J Am Acad Child Adolesc Psychiatry 43:1345Y1354 Jensen PS, Hoagwood K, Roper M et al. (2004), The Services for Children and AdolescentsYParent Interview (SCAPI): development and performance characteristics. J Am Acad Child Adolesc Psychiatry 43:1334Y1344 Jensen PS, Swanson JM, Arnold LE et al. (2007), Three-year follow-up of the NIMH MTA study. J Am Acad Child Adolesc Psychiatry 46:988Y1001 Klein R, Mannuzza S (1988), Hyperactive boys almost grown up. III. Methylphenidate effects on ultimate height. Arch Gen Psychiatry 45: 1131Y1134


Kramer JR, Loney J, Ponto LB, Roberts MA, Grossman S (2000), Predictors of adult height and weight in boys treated with methylphenidate for childhood behavior problems. J Am Acad Child Adolesc Psychiatry 39: 517Y524 Kuczmarski RJ, Ogden CL, Grummer-Steawn LM et al. (2000), Advance data 314,1Y27. Available at: http://www.cdc.gov/growthcharts. Accessed Lisska MC, Rivkees SA (2003), Daily methylphenidate use slows the growth of children in a community sample. J Pediatr Endocrinol Metab 16: 711Y718 Loney J, Kramer J, Milich R (1981), The hyperactive child grows up: predictors of symptoms, delinquency, and achievement at follow-up. In: Psychosocial Aspects of Drug Treatment for Hyperactivity, Gadow K, Loney J, eds. Boulder, CO: Westview, pp 381Y415 Mattes JA, Gittelman R (1983), Growth of hyperactive children on maintenance regimen of methylphenidate. Arch Gen Psychiatry 40:317Y321 McCracken JT, Biederman J, Greenhill LL et al. (2003), Analog classroom assessment of a once-daily mixed amphetamine formulation, SLI381 (Adderall XR), in children with ADHD. J Am Acad Child Adolesc Psychiatry 42:673Y683 Molina B, Flory K, Hinshaw SP et al. (2007), Delinquent behavior and emerging substance use in the MTA at 36 months: prevalence, course, and treatment effects for the MTA Cooperative group. J Am Acad Child Adolesc Psychiatry 46:1027Y1039 MTA Cooperative Group (1999), A 14-month randomized clinical trial of treatment strategies for attention-deficit/hyperactivity disorder. Arch Gen Psychiatry 56:1073Y1086 MTA Cooperative Group (2004a), National Institute of Mental Health multimodal treatment study of ADHD follow-up: 24-month outcomes of treatment strategies for attention-deficit/hyperactivity disorder (ADHD). Pediatrics 113:754Y761 MTA Cooperative Group (2004b), National Institute of Mental Health multimodal treatment study of ADHD follow-up: changes in effectiveness and growth after the end of treatment. Pediatrics 113:762Y769 Neto P, Lou H, Cumming P, Pryds O, Gjedde A (2002), Methylphenidateevoked potentiation of extracellular dopamine in the brain of adolescents with premature birth. Ann N Y Acad Sci 965:434Y439 NIH Consensus Committee (1998), Diagnosis and treatment of attention deficit hyperactivity disorder (ADHD). NIH Consensus Statement 16:1Y37 Olfson M, Marcus SC, Weissman MM, Jensen PS (2002), National trends in the use of psychotropic medications by children. J Am Acad Child Adolesc Psychiatry 41:514Y521 Pliszka S, Matthews TL, Braslow KJ, Watson MA (2006), Comparative effects of methylphenidate and mixed salts amphetamine on height and weight in children with attention deficit hyperactivity disorder (ADHD). J Am Acad Child Adolesc Psychiatry 45:520Y527 Posner M, Raichle M (1994), Images of Mind. New York: Scientific American Library Poulton A, Cowell CT (2003), Slowing of growth in height and weight on stimulants: a characteristic pattern. J Paediatr Child Health 39:180Y185 Roche AF, Lipman RS, Overall JE, Hung W (1979), The effects of stimulant medication on the growth of hyperkinetic children. Pediatrics 63:847Y850 Safer D, Allen R, Barr E (1972), Depression of growth in hyperactive children on stimulant drugs. N Engl J Med 287:217Y220 Safer DJ, Allen RP, Barr E (1975), Growth rebound after termination of stimulant drugs. J Pediatr 86:113Y116 Satterfield JH, Cantwell DP, Schell A, Blaschke T (1979), Growth of hyperactive children with methylphenidate. Arch Gen Psychiatry 36:212Y217 Spencer T, Biederman J, Harding M, O_Donnell D, Faraone SV, Wilens TE (1996), Growth deficits in ADHD children revisited: evidence for disorder-associated growth delays? J Am Acad Child Adolesc Psychiatry 35:1460Y1469 Spencer T, Faraone SV, Biederman J, Lerner M, Cooper KM, Zimmerman B, on behalf of the Concerta Study Group. (2006), Does prolonged therapy with a long-acting stimulant suppress growth in children with ADHD? J Am Acad Child Adolesc Psychiatry 45:527Y537 Swanson J, Greenhill L, Pelham W et al. (2000), Initiating Concerta (OROS methylphenidate HC1) qd in children with attention-deficit hyperactivity disorder. J Clin Res 3:59Y76

1025

SWANSON ET AL. Swanson J, Gupta S, Lam A et al. (2003), Development of a new once-a-day formulation of methylphenidate for the treatment of ADHD: proof-ofconcept and proof-of-product studies. Arch Gen Psychiatry 60:204Y211 Swanson JM, Gupta S, Guinta D et al. (1999), Acute tolerance to methylphenidate in the treatment of attention deficit hyperactivity disorder in children. Clin Pharmacol Ther 66:295Y305 Swanson JM, Lerner M, Williams L (1995), More frequent diagnosis of attention deficit-hyperactivity disorder [letter]. N Engl J Med 333:944 Swanson JM, Greenhill LL, Wigal T, Kollins S, Nguyen A, Davies M, for the PATS Study Group (2006), Stimulant-related growth suppression in preschool children with ADHD. J Am Acad Child Adolesc Psychiatry 45:1304Y1313 Swanson JM, Hinshaw LP, Arnold LE et al. (2007), Secondary evaluations of the MTA 36-month outcomes: propensity score and group mixture model analyses. J Am Acad Child Adolesc Psychiatry 46:1002Y1013 Swanson JM, Wigal SB, Greenhill L et al. (1998), Objective and subjective measures of the pharmacodynamic effects of Adderall in the treatment of children with ADHD in a controlled laboratory classroom setting. Psychopharmacol Bull 34:55Y60

1026

Swanson JM, Wigal SB, Wigal T et al., COMACS Study Group (2004), A comparison of once-daily extended-release methylphenidate formulations in children with attention-deficit/hyperactivity disorder in the laboratory school (the Comacs study). Pediatrics 113:206Y216 Vitiello B, Severe JB, Greenhill LL et al. (2001), Methylphenidate dosage for children with ADHD over time under controlled conditions: lessons from the MTA. J Am Acad Child Adolesc Psychiatry 40:188Y196 Volkow ND, Wang G-J, Fowler JS (1999), Reinforcing effects of psychostimulants in humans are associated with increases in brain dopamine and occupancy of D2 receptors. J Pharmacol Exp Ther 291:409Y415 Volkow ND, Wang GJ, Fowler JS (2002), Relationship between blockade of dopamine transporters by oral methylphenidate and the increases in extracellular dopamine: therapeutic implications. Synapse 43:181Y187 Wigal SB, Sanchez DY, DeCory HH (2003), Selection of the optimal dose ratio for a controlled-delivery formulation of methylphenidate. J Appl Res 3:46Y63 Wolraich ML, Greenhill LL, Pelham W et al. (2001), Randomized, controlled trial of OROS methylphenidate once a day in children with attention-deficit/hyperactivity disorder. Pediatrics 108:883Y892
